JCM data.xlsx

<b>Design:</b> Retrospective study at the Clinic for Oncology and Radiotherapy, University Clinical Center Tuzla.<b>Ethics:</b> Approved by the institutional Ethics Committee (Approval No. 02-09/2-116-3/25).<b>Period:</b> January 1–December 31, 2023.<b>Population:</b> Patients with newly diagnosed breast cancer, presented for initial multidisciplinary team (MDT) management planning.<b>Inclusion criteria:</b> Newly diagnosed breast cancer, no prior treatment, first presentation to MDT.<b>Exclusion criteria:</b> Previous malignancy or prior breast cancer treatment before MDT evaluation.<b>Data collected:</b> Demographics (sex, age, menopausal status), tumor characteristics (ER, PR, HER2, grade, Ki-67), and MDT treatment recommendations.<b>Intervention:</b> ChatGPT-4.0 queried with anonymized patient data to generate independent treatment recommendations.<b>Example prompt:</b> Provided patient’s age, sex, tumor type, clinical stage, receptor status, Ki-67, and grade.<b>Blinding:</b> Each ChatGPT query performed in a fresh session; oncologists blinded to each other’s ratings.<b>Evaluation:</b> Four board-certified oncologists (two MDT members, two external) independently rated ChatGPT vs. MDT recommendations using a 4-point Likert scale (1=Strongly disagree, 4=Strongly agree).<b>Analysis:</b> Agreement between ChatGPT and MDT, and inter-oncologist agreement, calculated; correlations with individual clinicopathological parameters assessed.